CN105426988A - Spacial load prediction method based on fuzzy rule - Google Patents

Abstract

A spatial load prediction method based on a fuzzy rule includes the steps of: S1. obtaining cell division information, various calculation definition tables and spatial data of cells of a region to be predicted, and dynamically generating a fuzzy rule table; S2. performing fuzzy evaluation on the cells through cell spatial data fuzzification, fuzzy inference and defuzzification calculation of the cells, obtaining a comprehensive evaluation table of the cells in developing lands of various types, and distributing areas for the cells to develop the lands of various types according to the comprehensive evaluation table; and S3. according to a load density fluctuation function of the lands of various types and the areas of the cells for developing the lands of various types, obtaining a load increase situation of the cells of the region to be predicted, and further obtaining a spatial load prediction result of the region to be predicted. The spatial load prediction method based on a fuzzy rule divides a complex inference rule into two levels to perform inference, thereby reducing the difficulty of load prediction, and lowering a requirement for precise data, and the method is relatively practical and has strong universality.

Description

A kind of Spatial Load Forecasting method based on fuzzy rule

Technical field

The present invention relates to a kind of Spatial Load Forecasting method based on fuzzy rule.

Background technology

Distribution network planning needs to determine the position of transformer station, the trend of capacity and feeder line and type, so during distribution network load prediction, not only will comprise the estimation to following load capacity, and comprises the prediction required load type, geographic distribution etc.The load prediction of power distribution network is the basis of distribution network planning, and for whole planning specifies target, the height of the accuracy of load prediction, will directly have influence on effect and the feasibility of planning.Distribution network planning relates to the relevant information in a large amount of geographic position, and the geographic distribution of load will directly have influence on the result of planning, therefore, in distribution network planning, essential to the prediction of space load.

The geographic distribution of load is predicted, generally following service area is divided into some regional ensembles, to predict each district's load growth situation.Most of Modern Regional Load Forecasting method mostly adopts trend extrapolation, namely with curve or additive method extrapolation historical load peak, but although trend extrapolation can obtain peak load density everywhere, but it is to data accuracy requirement is higher, and precision of prediction is lower, and due to actual conditions such as land used purposes, urban renewals, land used rule is often difficult to determine, for physical planning brings difficulty.

Summary of the invention

The object of the invention is for the deficiencies in the prior art, propose a kind of Spatial Load Forecasting method based on Fuzzy rule evaluation adopting the demand reduced precise information, the actual conditions that can meet planning at present, be more suitable for modern power systems planning.

The present invention is achieved through the following technical solutions:

Based on a Spatial Load Forecasting method for Fuzzy rule evaluation, it is characterized in that: comprise the steps:

S1: obtain the community division information in region to be predicted, all kinds of calculating definition list, each community spatial data, dynamically generate fuzzy reasoning table;

S2: fuzzy evaluation is carried out to each community by calculating data obfuscation, fuzzy reasoning and anti fuzzy method between the cell null of each community respectively, obtain the comprehensive evaluation form that all types of land used is developed in each community, and pro rata distribute each community according to this comprehensive evaluation form and develop the area of all types of land used;

S3: the area developing all types of land used according to the load density of all types of land used change function and each community, the load growth situation of this each community, region to be measured can be drawn, the Spatial Load Forecasting result in region to be predicted after the load growth situation of each community being added up, can be drawn.

Further, described all kinds of calculating definition list comprises the load density definition list of the land-use style definition list of each community, space attribute definition list, subordinate function definition list, all types of land used;

Described land-use style definition in tabledefine residential estate, commercial land, industrial land three class land-use style;

Described space attribute definition list comprises the large class table of space attribute and space attribute group table, the large class of described space attribute large class table definition space attribute and each large class relative to the weight coefficient of all types of land used of development, the relation of space attribute group, each space attribute group and the large class of space attribute, the weight coefficient of each space attribute group that the described space attribute group table definition large class of each space attribute comprises.

Described subordinate function definition list is according to actual conditions definition subordinate function form;

The load density definition list of described all types of land used defines the load density change function of all types of land used, and described load density change function is such as formula shown in (1):

$\frac{dS\left(t\right)}{dt}=\mathrm{\α}S\left(t\right)+\mathrm{\β}---\left(1\right)$

Wherein, α with β is that S (t) represents load density function with prediction time and land-use style relevant coefficient, and t represents the prediction time, and this function representation load density over time.

Further, described fuzzy reasoning table is defined as with space attribute group as condition, to develop the form of fitness for result of certain land-use style, described fuzzy reasoning table is set up by the knowledge and experience of general knowledge, general rule, investigation result, planning personnel, and can carry out suitable adjustment according to the change of actual conditions.

Further, described fuzzy evaluation is secondary fuzzy evaluation, first described secondary fuzzy evaluation is evaluated according to the fitness treating estimation range under each space attribute group factor comprised in the large class of certain space attribute and develop certain land-use style, then draws the fitness evaluation of regional development land-use style to be predicted under this space attribute large class factor according to described evaluation result.

Further, described step S1 comprises the steps:

S11: Image Via Gis obtains the community division information in region to be measured, Region dividing to be measured is n community, enters step S12;

S12: judge whether according to actual needs to select typical module, if, enter step S13, described typical module refers to that the coding rule by each calculating definition list is numbered in order is stored in typical template storehouse, when needing to carry out Spatial Load Forecasting, can directly find from typical template storehouse and derive corresponding calculating definition list, if not, then select self-defined pattern, enter step S14, when described self-defined pattern refers to carry out spatial prediction at every turn, all new definition is carried out to all kinds of calculating definition list;

S13: derive all kinds of calculating definition list from typical template storehouse;

S14: Image Via Gis and the self-defined all kinds of calculating definition list of actual conditions, and all kinds of calculating definition lists defined are loaded in typical template storehouse, enter step S15;

S15: according to land-use style definition list and the space attribute group definition list in described region to be predicted, dynamically generate fuzzy reasoning table;

S16: utilize Geographic Information System to obtain the spatial data of each community, span tables of data;

Wherein, n is integer and n >=1.

Further, described step S2 comprises the steps:

S21: judge that whether all communities are complete all as calculated, if so, enter step S34, if not, enter step S22;

S22: define according to subordinate function in tablethe subordinate function of definition is by the spatial data of the i-th community in tablespace data fuzzy;

S23: adopt Mamdani reasoning principle, in conjunction with fuzzy reasoning table and secondary fuzzy evaluation, show that fuzzy rule set G (x) of certain land-use style is developed in the i-th community, described fuzzy rule set G (x) is using space attribute group as Inference Conditions, and the fitness of certain land-use style is developed as the reasoning results in the i-th community;

S24: in conjunction with subordinate function definition list and fuzzy rule set G (x), adopt such as formula the anti fuzzy method formula shown in (2), anti fuzzy method is carried out to the obfuscation data of the i-th community, obtains the fitness evaluation score value that all types of land used is developed in the i-th community under every bar fuzzy rule:

${\mathrm{\μ}}_{{\mathrm{SP}}^{*}}\left(z\right)=max\left\{min\right({\mathrm{\μ}}_{SP}\left(Z\right)\left)\right\}---\left(2\right)$

Wherein, Z={ ε ₁, ε ₂, Λ ε _m, ε ₁, ε ₂, Λ ε _mbe respectively first space attribute group, second space attribute group ..., a m space attribute group adjusts the distance membership function mui _dis(Z) degree of membership, μ _sP(Z), represent the subordinate function of different fitness, m is integer and m>=1;

S25: the i-th community, under the large class factor of each space attribute, calculates according to each bar fuzzy rule maximal value be the fitness evaluation score value developing each land-use style, the weight coefficient of this evaluation score value and the relatively each land-use style of the large class of each space attribute is inserted attribute evaluation in table;

S26: according to the formula shown in formula (3), calculate the result of step S26, obtains the comprehensive evaluation form of all types of lands used of the i-th community, described comprehensive evaluation in tablecomprise the comprehensive grading value that each land-use style is developed in the i-th community:

$P=\frac{\Σrx}{\Σr}---\left(3\right)$

Wherein, P is the comprehensive grading value of certain land-use style of development, and r is the weight coefficient of the large class of space attribute certain land-use style relative, and x is under the large class factor of space attribute, develops the fitness evaluation score value of certain land-use style;

S27: according to described comprehensive evaluation in tablethe comprehensive grading value of each land-use style, pro rata distributing the i-th community develops the area of all types of land used;

Wherein, i is cell number, 1≤i≤n.

Further, step S3 comprises the following steps:

S31: obtain the load density initial value S0 in first prediction time, the factor alpha of load density change function and the numerical value of β under the difference prediction time from statistical department;

S32: to first prediction time, the area i-th community being developed all types of land used is multiplied by load density change function, integration is carried out to multiplied result, S0 got by initial value, the value that factor alpha and β are corresponding under getting first prediction time, and using the initial value of integral result as the load density in next one prediction time;

S33: repeat step S32, until all prediction times have all calculated, namely obtained the load variations situation of the i-th community, enter step S21;

S34: the load growth situation of each community added up, can draw the Spatial Load Forecasting result in region to be predicted.

The present invention has following beneficial effect:

1, the present invention is in load prediction process, in conjunction with Fuzzy rule evaluation, when reducing the demand to precise information, can meet the actual conditions of planning at present, being more suitable for the planning of modern power systems.

2, the present invention dynamically generates fuzzy reasoning table according to regional space attribute to be predicted and actual land-use style, more realistic and highly versatile;

3, the present invention is in computation process, adopts the method for secondary fuzzy evaluation, is divided into secondary to carry out reasoning comparatively intricate reasoning rule, reduces the complexity that every one-level is evaluated, reduces the difficulty of load prediction;

4, the present invention is when dynamically generating fuzzy reasoning table, generates an independent fuzzy reasoning table and carry out name according to certain rule to preserve for the large class of each space attribute, reduces the complicacy of calculating, improves efficiency.

Accompanying drawing explanation

Below in conjunction with accompanying drawingthe present invention is described in further details.

fig. 1for steps flow chart of the present invention figure.

fig. 2for the distance subordinate function in one embodiment of the invention.

fig. 3for the fitness subordinate function in one embodiment of the invention.

fig. 4for the Spatial Load Forecasting result of A community in one embodiment of the invention.

Embodiment

as Fig. 1be depicted as steps flow chart of the present invention figurein the present invention, first the community division information in region to be predicted is obtained from Geographic Information System, region to be predicted is divided into n community, respectively Spatial Load Forecasting is carried out to each community again, predicting the outcome of each community is added up, be the Spatial Load Forecasting result in region to be predicted, wherein, n is integer and n>=1.

The step that Spatial Load Forecasting is carried out in each community mainly comprises information, land use decision-making and load prediction three steps, tell information comprise obtain region to be predicted community division information, obtain all kinds of calculating definition list, dynamically generate fuzzy reasoning table, obtain each community spatial data; Land use decision-making of telling, calculated by data obfuscation, fuzzy reasoning and anti fuzzy method between the cell null to each community and fuzzy evaluation is carried out to each community, obtain the comprehensive evaluation form that all types of land used is developed in each community, and pro rata distribute each community according to this comprehensive evaluation form and develop the area of all types of land used; Tell load prediction develops all types of land used area according to the load density of all types of land used change function and each community, the load growth situation of this each community, region to be measured can be drawn, and then draw the Spatial Load Forecasting result in region to be predicted.

The implementation step of carrying out Spatial Load Forecasting due to each community is all identical, so be only described in detail method of the present invention for one of them community in the present embodiment.

In the present embodiment, be described for A community, A community usable area is 1.8km ², realize the Spatial Load Forecasting to A community by following steps.

S11: Image Via Gis obtains the community division information in region to be predicted, A community is one of them community of Region dividing to be predicted;

S12: the present embodiment selects self-defined pattern, enters step S14, when described self-defined pattern refers to carry out spatial prediction at every turn, carries out new definition to all kinds of calculating definition list;

S14: Image Via Gis and the self-defined all kinds of calculating definition list of actual conditions, and all kinds of calculating definition lists defined are loaded in typical template storehouse, enter step S15;

Wherein, all kinds of calculating definition lists related in step S13, step S14 comprise the load density definition list of the land-use style definition list of A community, space attribute definition list, subordinate function definition list, all types of land used.Now respectively the particular content of all kinds of calculating definition lists in the present embodiment is described:

Described land-use style definition list can obtain from statistical department, particular content as table 1shown in, in the present embodiment, only pay close attention to " large class " content:

Large class	Group
		Residential estate	Residential building, villa quarter, school district
Commercial land	Retail trade, show business, office building, skyscraper
		Industrial land	Big-and-middle-sized industry, storage, factory building

table 1

Described space attribute definition list comprises the large class definition list of space attribute and space attribute group definition list, the large class of described space attribute large class table definition space attribute and each large class are to the weight coefficient of each land-use style of development, described weight coefficient for representing that this generic attribute develops the influence degree of certain land-use style to A community, the space attribute group definition that the described space attribute group table definition large class of each space attribute comprises, each space attribute group and the relation of the large class of space attribute, the weight coefficient of each space attribute group.

The weight coefficient of the large class of space attribute and the relatively each land-use style of each large class as table 2shown in:

table 2

Space attribute group table as table 3shown in, in the present embodiment, do not consider the weight coefficient of space attribute group:

Group	Affiliated large class
		Apart from city's centre distance	Apart from the distance at center
Apart from district center distance	Apart from the distance at center
		Apart from arterial traffic distance	Geographic position
Apart from nearest school distance	Geographic position
		Apart from nearest market distance	Geographic position
Apart from neighbourhood offset from	Environmental factor
		Apart from adjacent commercial offset from	Environmental factor
Apart from adjacent manufacturing district distance	Environmental factor

table 3

Described subordinate function definition list defines subordinate function form, the semanteme of certain attribute ambiguity is showed rightly according to actual conditions.In the present embodiment, for distance, define three fuzzy sets " closely ", concept that " medium ", " far " represent distance, distance subordinate function as Fig. 2shown in, wherein " C " expression " closely ", " M " expression " medium ", " F " expression " far ", when subordinate function is 1, represent that certain distance meets the implication expressed by concept completely, when subordinate function is 0, represent that certain distance does not meet the implication expressed by concept completely. fig. 2in have the cross section of 20%-50% between three subordinate functions, show the ambiguity of concept, such as the distance of 1.5km, represent that apart from the subjection degree of " closely " be 0.25, represent that apart from the subjection degree of " medium " be 0.75, namely this distance both can be referred to as " closely ", also can be referred to as " medium ".

The membership function representing all types of land used fitness is also defined in the present embodiment, described each land-use style fitness represents that the suitable degree of certain type land used is developed in A community under a certain space attribute factor, namely A community is suitable for the degree developing certain type land used,: SA represents and is far from suitable to define five fuzzy sets to represent fitness concept, MA represents more inadaptable, and NT represents general, and MP represents and comparatively adapts to, SP represents and adapts to very much, concrete five subordinate functions as Fig. 3shown in, respectively there is partial intersection between each subordinate function, embody the ambiguity of fitness, such as, for fitness x=0.4, the corresponding μ of its membership function value _mA=0.5, corresponding μ _nT=0.5, represent that this fitness x contains the implication of MA, the implication again containing NT.But x does not have the implication of SA, MP, SP, because now the membership function value of these three concepts is expressed as 0.

The load density definition list of described all types of land used defines the load density change function of all types of land used, and described load density change function is shown below:

Wherein, α with β is that S (t) represents load density function with prediction time and land-use style relevant coefficient, and t represents the prediction time, and this function representation load density over time.The load total amount of all types of land used is increased statistics and can be obtained by time load prediction system, and the coefficient value of load density change function can be obtained by statistical department, each coefficient value in the present embodiment as table 4shown in:

table 4

table 4in, C is the constant representing the prediction time, S ₀represent the initial value used when the load prediction in calculating first prediction time, the initial value of use in each year is afterwards the result calculated for a year.

S15: according to land-use style definition list and the space attribute group definition list of described A community, dynamically generate fuzzy reasoning table;

In order to make fuzzy reasoning table easy to use, conveniently carry out secondary fuzzy evaluation, when dynamically generating fuzzy reasoning table, in conjunction with the define method of space attribute definition list, generate an independent fuzzy reasoning table for the large class of each space attribute and name according to certain rule, like this, when such changes, only need regenerate corresponding fuzzy reasoning table, and do not need the fuzzy reasoning table having influence on other, relative to generating a large fuzzy reasoning table for the large class of all space attributes, greatly service efficiency can be improved.And when carrying out fuzzy reasoning, the number of condition can be reduced, effectively reduce the complexity of reasoning.Such as " apart from center " the large class of this space attribute, comprise " apart from city's centre distance " and " apart from district center distance " two groups, for the fuzzy reasoning table that the large class of this space attribute generates as table 5shown in:

Rule	Apart from city's centre distance	Apart from district center distance	Inhabitation fitness	Business fitness	Industry fitness
						1
2
						3

table 5

table 5in, " apart from city centre distance ", " apart from district center distance " as the condition of fuzzy rule, the result that " inhabitation fitness ", " business fitness ", " industrial fitness work " are fuzzy rule, fuzzy rule is:

If is apart from city centre distance AND apart from district center distance then land-use style fitness

When " geographic position " is as the condition of fuzzy rule, the result that " inhabitation fitness ", " business fitness ", " industrial fitness work " are fuzzy rule, can generate by following fuzzy rule as table 5shown fuzzy reasoning table:

If is apart from arterial traffic distance AND apart from nearest school distance AND apart from nearest market distance then land-use style fitness

When " environmental factor " is as the condition of fuzzy rule, the result that " inhabitation fitness ", " business fitness ", " industrial fitness work " are fuzzy rule, can generate by following fuzzy rule as table 5shown fuzzy reasoning table:

If is apart from neighbourhood offset from AND apart from adjacent commercial offset from AND apart from adjacent manufacturing district distance then land-use style fitness

In actual applications, according to the type of space attribute in certain cell data obtained, right as table 5shown fuzzy rule in tablea few rules trigger, thus form the fuzzy reasoning table adapted with A community.

S16: utilize Geographic Information System to obtain the spatial data of A community, span tables of data as table 6shown in, enter step S23;

Region	Space attribute group	The large class of space attribute	Data
				A	Apart from city's centre distance	Apart from the distance at center	3.5km
A	Apart from district center distance	Apart from the distance at center	2km
				A	Apart from arterial traffic distance	Geographic position	1km
A	Apart from nearest school distance	Geographic position	0.7km
				A	Apart from nearest market distance	Geographic position	0.4km
A	Apart from neighbourhood offset from	Environmental factor	0.4km
				A	Apart from adjacent commercial offset from	Environmental factor	1.2km
A	Apart from adjacent manufacturing district distance	Environmental factor	8km

table 6

S22: according to distance membership function μ _dis(Z) ( fig. 2) by the space data fuzzy of A community:

Space attribute group	The large class of space attribute	Data	Closely (C)	Medium (M)	Far (F)
						Apart from city's centre distance	Apart from the distance at center	3.5km	0	1	0
Apart from district center distance	Apart from the distance at center	2km	0	1	0
						Apart from nearest market distance	Geographic position	0.2km	0.8	0.2	0
Apart from main line of communication distance	Geographic position	1km	0.5	0.5	0
						Apart from nearest school distance	Geographic position	0.7km	0.65	0.35	0
Apart from neighbourhood offset from	Environmental factor	0.4km	0.8	0.2	0
						Apart from adjacent commercial offset from	Environmental factor	1.2km	0.4	0.6	0
Apart from adjacent manufacturing district distance	Environmental factor	8km	0	0	1

table 7

S23: adopt Mamdani reasoning principle, in conjunction with as table 5shown fuzzy reasoning table and corresponding fuzzy rule, secondary fuzzy evaluation, show that fuzzy rule set G (x) of certain land-use style is developed in A community, fuzzy rule in described fuzzy rule set G (x) is using space attribute group as Inference Conditions, and A develops community the fitness of certain land-use style as the reasoning results;

In the present embodiment, according to fuzzy rule mentioned above, for development inhabitation type land used, the fuzzy reasoning table of the large class of space attribute " distance apart from center " as table 8shown in-1; The fuzzy reasoning table of the large class of space attribute " geographic position " as table 8shown in-2; The fuzzy reasoning table of the large class of space attribute " environmental factor " as table 8shown in-3:

Rule	Apart from city's centre distance	Apart from district center distance	Inhabitation fitness
				1-1	Far (F)	Closely (C)	Be far from suitable (SA)
1-2	Far (F)	Far (F)	Comparatively inadaptable (MA)
				1-3	Medium (M)	Far (F)	Generally (NT)
1-4	Closely (C)	Closely (C)	Comparatively adapt to (MP)

1-5

Medium (M)

Medium (M)

Adapt to (SP) very much

table 8-1

Rule	Apart from nearest market distance	Apart from main line of communication distance	Apart from nearest school distance	Inhabitation fitness
					2-1	Far (F)	Closely (C)	Far (F)	Be far from suitable (SA)
2-2	Far (F)	Closely (C)	Closely (C)	Comparatively inadaptable (MA)
					2-3	Medium (M)	Medium (M)	Medium (M)	Generally (NT)
2-4	Closely (C)	Far (F)	Medium (M)	Comparatively adapt to (MP)
					2-5	Closely (C)	Medium (M)	Closely (C)	Adapt to (SP) very much

table 8-2

Rule	Apart from neighbourhood offset from	Apart from adjacent commercial offset from	Apart from adjacent manufacturing district distance	Inhabitation fitness
					3-1	Far (F)	Far (F)	Closely (C)	Be far from suitable (SA)
3-2	Far (F)	Far (F)	Medium (M)	Comparatively inadaptable (MA)
					3-3	Medium (M)	Medium (M)	Medium (M)	Generally (NT)
3-4	Closely (C)	Medium (M)	Medium (M)	Comparatively adapt to (MP)
					3-5	Closely (C)	Medium (M)	Far (F)	Adapt to (SP) very much

table 8-3

S24: in conjunction with G (x) and fitness subordinate function ( fig. 3shown in), according to formula anti fuzzy method is carried out to the obfuscation data of A community, obtains the fitness evaluation score value that all types of land used is developed in A community under each fuzzy rule, as table 9shown in:

Wherein, Z={ ε ₁, ε ₂, Λ ε _m, ε ₁, ε ₂, Λ ε _mbe respectively first space attribute group, second space attribute group ..., a m space attribute group adjusts the distance membership function mui _dis(Z) degree of membership, as table 7in " near (C) ", " medium (M) ", value corresponding to " (F) far away ", G (x) represents the set of all fuzzy rules, according to secondary fuzzy evaluation, by all fuzzy rules spatially the large class of attribute be divided three classes, respectively as table8-1, table 8-2, table 8shown in-3, μ _sP(Z) fitness subordinate function is represented, as Fig. 3shown in, represent that the fitness evaluation score value of all types of land used is developed in A community under each fuzzy rule, wherein, " degree of membership 1 ", " degree of membership 2 ", " degree of membership 3 " correspondence table 7in " near (C) ", " medium (M) ", " (F) far away ", " fitness 1 ", " fitness 2 ", " fitness 3 " are for bring into " degree of membership 1 ", " degree of membership 2 ", " degree of membership 3 " as Fig. 3shown membership function mui _sP, and in conjunction with the numerical value of fuzzy rule gained, the acquisition of this numerical value have employed minimum membership degree method, namely gets the minimum value in fitness interval corresponding to degree of membership, and " regular fitness " is (Z) the minimum value of each fitness under being taken at each rule:

Number of regulation	Degree of membership 1	Fitness 1	Degree of membership 2	Fitness 2	Degree of membership 3	Fitness 3	Rule fitness
								1-1	0	0	0	0	/	/	0
1-2	0	0	0	0	/	/	0
								1-3	1	0.5	0	0.3	/	/	0.3
1-4	0	0.5	0	0.5	/	/	0.5
								1-5	1	1	1	1	/	/	1
2-1	0	0	0.5	0.1	0	0	0
								2-2	0	0	0.5	0.1	0.65	0.13	0
2-3	0.2	0.34	0.5	0.4	0.35	0.37	0.34
								2-4	0.8	0.66	0	0.5	0.35	0.57	0.5
2-5	0.8	0.96	0.5	0.9	0.65	0.93	0.9
								3-1	0	0	0	0	0	0	0
3-2	0	0	0	0	0	0	0
								3-3	0.2	0.34	0.6	0.42	0	0.3	0.3
3-4	0.8	0.66	0.6	0.62	0	0.5	0.5
								3-5	0.8	0.96	0.6	0.92	1	1	0.92

table 9

Obtain the step of the large class of each space attribute of A community to the evaluation score value of the fitness of development business type land used, industrial type land used identical with above-mentioned steps.

S25: under the large class factor of each space attribute, calculate according to each bar fuzzy rule maximal value be the fitness evaluation score value that each land-use style is developed in A community, the weight of this evaluation score value and the large class of each space attribute certain land-use style is relatively inserted attribute evaluation in table, as table 1shown in 0:

table 10

S26: according to the formula shown in formula (3), calculate the result of step S26, obtains the comprehensive evaluation form of all types of lands used of A community, and described comprehensive evaluation form comprises the comprehensive grading value that each land-use style is developed in A community, as table 1shown in 1:

$P=\frac{\Σrx}{\Σr}---\left(3\right)$

Wherein, P is the comprehensive grading value of certain land-use style of development, and r is the weight coefficient of the large class of space attribute certain land-use style relative, and x is under the large class factor of space attribute, develops the fitness evaluation score value of certain land-use style;

S27: according to table 11, by the area of all types of land used in proportional distribution A community shared by comprehensive grading value of each land-use style of development, enter step S31:

Residential estate area is: $1.8\×\left(\frac{0.9393}{0.9393+0.2572+0.0539}\right)=1.3522{\mathrm{km}}^2$

Commercial land area is: $1.8\×\left(\frac{0.2572}{0.9393+0.2572+0.0539}\right)=0.3702{\mathrm{km}}^2$

Industrial land area is: $1.8\×\left(\frac{0.0539}{0.9393+0.2572+0.0539}\right)=0.0776{\mathrm{km}}^2$

S31: obtain load density function at the numerical value of first the prediction load density initial value S0 in time, factor alpha and β under the difference prediction time from statistical department, as table 4shown in;

S32: to first prediction time, is multiplied by load density change function by the area of all types of for A community land used carry out integration to multiplied result, S0 got by initial value, the value that factor alpha and β are corresponding under getting first prediction time, and using the initial value of integral result as the load density in next one prediction time;

S33: repeat step S32, until all prediction times have all calculated, namely obtain the load growth situation of A community, as Fig. 4shown in, wherein, for each prediction time, the load sum of each land-use style should equal total load amount.

Above-mentioned for A community, Spatial Load Forecasting method is described, in actual applications, for each community of Region dividing to be predicted, said method is all adopted to carry out Spatial Load Forecasting, then predicting the outcome of each community is added up, the space load result of community to be predicted can be obtained.

The above, be only preferred embodiment of the present invention, therefore can not limit scope of the invention process with this, and the equivalence namely done according to the present patent application the scope of the claims and description changes and modifies, and all should still remain within the scope of the patent.

Claims (7)

1., based on a Spatial Load Forecasting method for Fuzzy rule evaluation, it is characterized in that: comprise the steps:

S1: obtain the community division information in region to be predicted, all kinds of calculating definition list, each community spatial data, dynamically generate fuzzy reasoning table;

S2: fuzzy evaluation is carried out to each community by calculating data obfuscation, fuzzy reasoning and anti fuzzy method between the cell null of each community respectively, obtain the comprehensive evaluation form that all types of land used is developed in each community, and pro rata distribute each community according to this comprehensive evaluation form and develop the area of all types of land used;

S3: the area developing all types of land used according to the load density of all types of land used change function and each community, the load growth situation of this each community, region to be measured can be drawn, the Spatial Load Forecasting result in region to be predicted after the load growth situation of each community being added up, can be drawn.

2. a kind of Spatial Load Forecasting method based on Fuzzy rule evaluation according to claim 1, is characterized in that: described all kinds of calculating definition list comprises the load density definition list of the land-use style definition list of each community, space attribute definition list, subordinate function definition list, all types of land used;

Residential estate, commercial land, industrial land three class land-use style is defined in described land-use style definition list;

Described space attribute definition list comprises the large class table of space attribute and space attribute group table, the large class of described space attribute large class table definition space attribute and each large class relative to the weight coefficient of all types of land used of development, the relation of space attribute group, each space attribute group and the large class of space attribute, the weight coefficient of each space attribute group that the described space attribute group table definition large class of each space attribute comprises.

Described subordinate function definition list is according to actual conditions definition subordinate function form;

The load density definition list of described all types of land used defines the load density change function of all types of land used, and described load density change function is such as formula shown in (1):

$\frac{dS\left(t\right)}{dt}=\mathrm{\α}S\left(t\right)+\mathrm{\β}---\left(1\right)$

Wherein, α with β is that S (t) represents load density function with prediction time and land-use style relevant coefficient, and t represents the prediction time, and this function representation load density over time.

3. a kind of Spatial Load Forecasting method based on Fuzzy rule evaluation according to claim 1, it is characterized in that: described fuzzy reasoning table is defined as with space attribute group as condition, to develop the form of fitness for result of certain land-use style, described fuzzy reasoning table is set up by the knowledge and experience of general knowledge, general rule, investigation result, planning personnel, and can carry out suitable adjustment according to the change of actual conditions.

4. a kind of Spatial Load Forecasting method based on Fuzzy rule evaluation according to claim 1 and 2, it is characterized in that: described fuzzy evaluation is secondary fuzzy evaluation, first described secondary fuzzy evaluation is evaluated according to the fitness treating estimation range under each space attribute group factor comprised in the large class of certain space attribute and develop certain land-use style, then draws the fitness evaluation of regional development land-use style to be predicted under this space attribute large class factor according to described evaluation result.

5. a kind of Spatial Load Forecasting method based on Fuzzy rule evaluation according to claim 1 and 2, is characterized in that: described step S1 comprises the steps:

S11: Image Via Gis obtains the community division information in region to be measured, Region dividing to be measured is n community, enters step S12;

S12: judge whether according to actual needs to select typical module, if, enter step S13, described typical module refers to that the coding rule by each calculating definition list is numbered in order is stored in typical template storehouse, when needing to carry out Spatial Load Forecasting, can directly find from typical template storehouse and derive corresponding calculating definition list, if not, then select self-defined pattern, enter step S14, when described self-defined pattern refers to carry out spatial prediction at every turn, all new definition is carried out to all kinds of calculating definition list;

S13: derive all kinds of calculating definition list from typical template storehouse;

S14: Image Via Gis and the self-defined all kinds of calculating definition list of actual conditions, and all kinds of calculating definition lists defined are loaded in typical template storehouse, enter step S15;

S15: according to land-use style definition list and the space attribute group definition list in described region to be predicted, dynamically generate fuzzy reasoning table;

S16: utilize Geographic Information System to obtain the spatial data of each community, span tables of data;

Wherein, n is integer and n >=1.

6. a kind of Spatial Load Forecasting method based on Fuzzy rule evaluation according to claim 1 or 2 or 3 or 4, is characterized in that: described step S2 comprises the steps:

S21: judge that whether all communities are complete all as calculated, if so, enter step S34, if not, enter step S22;

S22: according to the subordinate function defined in subordinate function definition list by the space data fuzzy in the spatial data table of the i-th community;

S23: adopt Mamdani reasoning principle, in conjunction with fuzzy reasoning table and secondary fuzzy evaluation, show that fuzzy rule set G (x) of certain land-use style is developed in the i-th community, described fuzzy rule set G (x) is using space attribute group as Inference Conditions, and the fitness of certain land-use style is developed as the reasoning results in the i-th community;

S24: in conjunction with subordinate function definition list and fuzzy rule set G (x), adopt such as formula the anti fuzzy method formula shown in (2), anti fuzzy method is carried out to the obfuscation data of the i-th community, obtains the fitness evaluation score value that all types of land used is developed in the i-th community under every bar fuzzy rule:

Wherein, Z={ ε ₁, ε ₂, Λ ε _m, ε ₁, ε ₂, Λ ε _mbe respectively first space attribute group, second space attribute group ..., a m space attribute group adjusts the distance membership function mui _dis(Z) degree of membership, μ _sP(Z), z () represents the subordinate function of different fitness, m is integer and m>=1;

S25: the i-th community, under the large class factor of each space attribute, calculates according to each bar fuzzy rule maximal value be the fitness evaluation score value developing each land-use style, the weight coefficient of this evaluation score value and the relatively each land-use style of the large class of each space attribute is inserted in attribute evaluation table;

S26: according to the formula shown in formula (3), calculate the result of step S26, obtains the comprehensive evaluation form of all types of lands used of the i-th community, and described comprehensive evaluation form comprises the comprehensive grading value that each land-use style is developed in the i-th community:

$P=\frac{\mathrm{\Σ}rx}{\mathrm{\Σ}r}---\left(3\right)$

Wherein, P is the comprehensive grading value of certain land-use style of development, and r is the weight coefficient of the large class of space attribute certain land-use style relative, and x is the fitness evaluation score value developing certain land-use style under the large class factor of space attribute;

S27: according to the comprehensive grading value of land-use style each in described comprehensive evaluation form, pro rata distributing the i-th community develops the area of all types of land used;

Wherein, i is cell number, 1≤i≤n.

7. a kind of Spatial Load Forecasting method based on Fuzzy rule evaluation according to claim 1 or 2 or 6, is characterized in that: step S3 comprises the following steps:

S31: obtain the load density initial value S0 in first prediction time, the factor alpha of load density change function and the numerical value of β under the difference prediction time from statistical department;

S32: to first prediction time, the area i-th community being developed all types of land used is multiplied by load density change function, integration is carried out to multiplied result, S0 got by initial value, the value that factor alpha and β are corresponding under getting first prediction time, and using the initial value of integral result as the load density in next one prediction time;

S33: repeat step S32, until all prediction times have all calculated, namely obtained the load variations situation of the i-th community, enter step S21;

S34: the load growth situation of each community added up, can draw the Spatial Load Forecasting result in region to be predicted.